A Practical Guide for Health Researchers - part 4 ppsx

Chapter 6 Submitting a research proposal 6.1 Introduction A research proposal is a document written for the purpose of obtaining funding for a research project. Researchers should familiarize themselves with the potential sources for funding, and their specific requirements and mechanisms. They should know how to submit a proposal that will have a good chance of getting funded. Grantsmanship is the term used for the ability to secure grants to support research projects. The research proposal includes all the components of the research protocol outlined in the previous chapter. In addition, the proposal has to include additional information to convince the funding agency that the project is worthy of support and can be successfully implemented. 6.2 How to get your research project funded 6.2.1 Sources of funding Funding for health research basically comes from either public sources or private sources. Public sources include governments and intergovernmental organizations. Private sources include the not-for-profit sector, such as philanthropic foundations and nongovernmental organizations, and the for-profit private industry. Besides these primary sources, there are intermediary agencies/organizations which play a role in channelling funding from the primary sources to the actors in research. Government funding is provided through publicly funded national research organizations, such as national research councils, institutes of health and universities. Some ministries of health see the value of health research for their work, and allocate a budget for it. Governments in developed countries may allocate funds for research through their bilateral official development assistance to developing countries. Two countries (Sweden and Canada) provide funding for research through publicly supported semi-autonomous agencies. The Swedish Agency for Research in Developing Countries (SAREC) and the International Development Research Centre (IDRC) in Canada provide a special mechanism for supporting research to solve developing country problems. This is trial version www.adultpdf.com Submitting a research proposal 73 Intergovernmental organizations, such as the World Health Organization, support research through provision of funding, as well as technical support. Support is provided through headquarters’ programmes, as well as through regional offices. Special research programmes in WHO cover the areas of reproductive health research and tropical disease research. The not-for-profit private sector includes several foundations, large and small. Examples include the Wellcome Trust in the UK, and the Rockefeller, Ford and Bill and Melinda Gates Foundations in the USA, among others. In the for-profit private sector, pharmaceutical companies, largely based in industrialized countries, are investing increasingly large sums of money in research and development. A layer of intermediary support often serves as a bridge between the funders and those conducting research. An increasing number of international research programmes have been active, focusing on particular areas of health research. Examples include the Population Council headquartered in New York and with a number of regional offices, Family Health International (FHI) headquartered in Chapel Hill (North Carolina) and the Program for Appropriate Technology in Health (PATH) with headquarters in Seattle. 6.2.2 Will the project be funded? Funding organizations receive many more proposals than what they can fund. The selection process is very competitive. The following factors are generally taken into consideration in deciding whether the proposal is to be funded: • importance and relevance of the research question to the declared interests of the agency; success in obtaining a grant depends on matching the proposal with the interests of a granting agency; • quality of the research design; • ability of the investigators to carry out the project; • capacity of the research facility to carry out the project; • ability of the institution to handle the administrative and financial procedures; • satisfactory ethical considerations; • realistic and justifiable budget, within the limits set by the agency, and normally with no expectations for continued funding after the completion of the project; • reasonable time-frame for completion of the project; • understanding of anticipated problems; • clarity and style of the written proposal. This is trial version www.adultpdf.com 74 A practical guide for health researchers Writing with enthusiasm is a good idea, but overstatements should be avoided. The applicant should be realistic about the limitations of the study. 6.2.3 How to submit a research proposal Funding organizations use one or more of the following mechanisms to select and fund research projects: solicit proposals, advertise and invite proposals or have an open door policy. • Soliciting proposals: In this case, one or more research institutions are approached and are asked for their interest in submitting a research proposal in a certain area of importance to the funding agency. Usually the institution approached is a centre of excellence. • Advertising an invitation for submitting proposals in certain areas of interest: Here, the process is competitive, and normally there is a time limit for submitting proposals. A part of grantsmanship is to be ready with good ideas, and to be able to speedily compile an attractive proposal. The proposals are then independently reviewed and scored, and a small number is selected for funding. Some funding agencies will ask first for a brief concept outline of the proposal, and then shortlist the applications, and ask for complete proposals from the short list. • An open door policy for any good proposal: Most funding agencies, however, have areas of interest and areas in which they are not interested. Good advice in this ap- proach is not to send a full proposal. A brief outline of the project with the level of funding requested is enough to get a response about the potential interest of the agency or organization in considering it. If the response is positive, the full proposal can be sent. Funding organizations have their own websites. Information about their interests and mechanisms can be easily accessed. 6.2.4 Response to comments of reviewers Research proposals are commonly subjected to peer review. The reviewers may suggest that the proposal can be made more acceptable by revisions. The investigators do not need to make all the changes suggested automatically. They should adopt revisions that will satisfy the reviewer’s criticisms wherever possible and justify any decision not to do so. It is good to indicate in a separate page the criticism made and how the revised proposal responded to them. This is trial version www.adultpdf.com Submitting a research proposal 75 6.3 Components of a research proposal A research proposal commonly follows the following format, which includes the components of the research protocol, with some additional information. Some funding agencies have their own formats for standardizing applications and streamlining the review process. • Title page • Project summary • Project description • Ethical considerations • Gender issues • Timetable • Problems anticipated • Budget • References • Curriculum vitae of the investigator(s) Title page This page should provide information on: • Project title • Principal investigator(s) • Institution • Duration of the project • Funding requested Research grants are normally given to institutions not to individuals. The name of the institution should be given in the title page. If the institution is not known to the agency, a brief information about the institution may be given as an annex, or may be requested by the agency. The name of the financial officer who will be in charge of administering the grant should be given, in addition to the names of the investigators. The duration of the project must be specified. Most agencies will not commit support beyond three years. The funding requested should be specified. In a multi-year project, the amount requested for each year should be outlined. This is trial version www.adultpdf.com 76 A practical guide for health researchers Project summary The project summary should be carefully written. It will be the first (and may be the only) part read by the reviewers. It should reveal persuasively the importance and the strengths of the project. Project description This should follow the lines of the protocol, as already discussed in the chapter on writing the protocol. The rationale should not only explain why the project is important to do, but should also indicate its relevance to the particular lines of interest of the funding agency. Previous work by the investigators on the research topic will indicate the competence of the investigators to carry out the work. Pilot studies, if already done, are important to demonstrate the feasibility of the research. Ethical considerations Approval from the local ethics review committee does not relieve the donor agency from the ethical responsibility for the project. Also approval by a donor agency does not relieve the research institution from ethical responsibility for the project. Ethical issues and concerns should be addressed fully in the research proposal, as outlined in the chapter on writing the research protocol. Gender issues Most funding organizations are now increasingly conscious about gender issues. These should be addressed in the proposal, as outlined in the previous chapter on writing the research protocol. Timetable The investigators should commit themselves to a timetable. This may include a preparatory phase to train research workers, to procure equipment/supplies, or to complete a pilot phase. The timetable should then estimate the duration for collection of data, final analysis of data and writing up the report. In project proposals of a long duration (more than one year), the timetable should set milestones to be reached. These are taken into consideration when progress reports are reviewed by the funding agency. Funding is often released on the basis of these progress reports. This is trial version www.adultpdf.com Submitting a research proposal 77 Problems anticipated The investigators should demonstrate their awareness of obstacles and difficulties, which may interfere with the successful completion of the project within the timeframe and cost proposed. They should explain how these obstacles and difficulties would be dealt with. An investigator who does not anticipate any problem probably has not thought out the details of the project carefully. Budget The budget request should be itemized and each item should be justified. Budget itemization The following are examples of categories of expenses: • Personnel (names, positions, percentage of time spent on the project, salary, fringe benefits) • Equipment • Supplies • Patient care costs • Travel • Data processing • Communications • Secretarial expenses • Publication/dissemination of information about the outcome of the project. Budget justification All items in the budget need to be justified and are closely scrutinized in the following way: • Are all personnel needed for the amount of time stated? • Are critical personnel devoting enough time to the project? • Major pieces of equipment are difficult to justify in a small project; an exception may be made for a developing country institution as part of research capability strengthening. • The budget should not include any undue inducement for subject participation. If the duration of the project is more than one year, a detailed budget is needed for at least the first year. Budget request for the subsequent years should be outlined. Agencies would normally approve the budget for the full duration of the project, but funds will be This is trial version www.adultpdf.com 78 A practical guide for health researchers released on a yearly basis, subject to the submission of acceptable progress and financial reports. Agencies normally will allow some flexibility in the use of the budget, provided the total budget is not exceeded. For shifts between budget items, however, it is expected that the agency’s approval be sought beforehand. An unrealistic budget is likely to lead to rejection of the proposal. The budget may be unrealistic in one of two ways. It may ask for more than is needed to undertake the project or it may ask for much less than is realistically needed to undertake the project successfully. The investigators may want to limit the budget to the funding ceiling of the agency, but keep the large project as it is. Instead, they should limit the project objectives to what can realistically be achieved with the requested funds. References A number of recent references on the subject should be cited in support of the proposal. Curriculum vitae (CVs) of the investigator(s) The ability of the investigators to carry out the project is an important consideration. Biographical sketches of the investigators or CVs should be attached. The track record of the investigators is important. Preliminary studies or other work done by the investigators on the subject should be included. References and additional sources of information Cummings SR, Holly EA, Hulley SB. Writing and funding a research proposal. In: Hulley SB, Cummings SR. eds. Designing clinical research: an epidemiologic approach, 2nd edition. Philadelphia, Lippincott Williams & Wilkins, 2001: 285–298. Wingo PA, Higgins JA, Rubin GL, Zahniser SC. An epidemiologic approach to reproductive health. Geneva. World Health Organization, 1994: 15–67 (WHO/HRP/EPI/1994). Health research methodology: A guide for training in research methods. 2nd edition. Manila, World Health Organization Regional office for the Western Pacific, 2001: 147–161. UNDP/UNFPA/WHO/World Bank Special Programme of Research, Development and Research Training in Human Reproduction. Preparing a research project proposal. Guidelines and forms. 3rd edition. Geneva, World Health Organization, 2000 (WHO/ HRP/PP/2000). Wyatt KM, Dimmock PW. Applying for a grant. In: O’Brien PMS, Pipkin FB. Introduction to research methodology for specialists and trainees. London, Royal College of Obstetricians and Gynaecologists Press, 1999: 201–209. This is trial version www.adultpdf.com Chapter 7 Implementing the research project 7.1 Introduction How should research be done? The answer to this question can be given in one word: well. Whatever the reason for the research, and whatever the kind of research, it should be done well and should conform to established standards of scientific methodology. It has been said that there is only one type of research: good research. Bad research does not deserve the name of research. It is not enough that the research question has been well conceived, the appropriate research design selected, and a detailed protocol well thought out and written. All these provide a good anatomy for the research. Physiology matters even more. The research should be implemented with scientific rigour. 7.2 Scientific rigour The English word “rigour” literally means “strictness”. In scientific research, the term rigour is used to imply that: • the study protocol is being adhered to; • the research is conducted in accordance with established ethical standards; • meticulous and detailed records of all observations are maintained; • methods of measurement are used in an objective way to provide valid and reliable results; • data are analysed and interpreted using appropriate statistical methods to assess the validity of the results and their generalizability; • the researchers continue to be well versed with the literature on the subject during the study; • results are presented in such a way that other investigators can re-analyse the data using the same processes and methods and reach the same conclusions, and other investigators can replicate the study to confirm or refute the findings. This is trial version www.adultpdf.com 80 A practical guide for health researchers 7.3 Pre-testing the protocol It is always wise to pre-test the protocol after developing it. This is particularly important in large and expensive studies. What appears to be a straightforward and problem-free protocol may prove to have logistic and practical problems in implementation. The pre-test is sometimes called a pilot study. Based on the outcome of the pilot study, the protocol may be modified before the study proper is implemented. The pilot study can help in determining whether the required number, as well as composition, of study subjects will be recruited within the time frame of the study. The size of the sample may need to be modified, or alternative approaches of recruitment may be explored. The pilot study can help in testing the methods of measurement. If the study relies on how records have been kept, these records may be checked for accuracy and completion before the study is started. If a questionnaire has been designed, this will need to be pre- tested to check that the questions are clear without any ambiguity and that the answers will be consistent. Modifications may have to be made for the final instrument. If the methodology involves a clinical or laboratory measurement, this has to be tested for inter-rater and intra-rater reliability, i.e. for consistency in results obtained by different workers and by the same worker at different times. The pilot study can also help in testing the system for data management. Entering and editing the data from the pilot study will show whether the system is working well. This includes designing the forms for recording measurements, choosing a computer, developing programmes for data entry, management and analysis; and planning dummy tabulations to assure that the appropriate variables are collected. 7.4 Monitoring of the study The study should be monitored. In large clinical trials, a monitor may be appointed with the responsibility of reporting on the progress of the trial and for verification of data. There are two components to monitoring: data management (record keeping and data handling) and data quality (quality assurance and quality control). Record-keeping and handling of data All steps involved in data management should be documented in order to allow step-by-step retrospective assessment of the quality of the data and the performance of the clinical trial (“the audit paper trail concept”). A basic aspect of the integrity of data is the safeguarding of “blinding” with regard to assignment of subjects to different treatments. Subject files and other supporting data must be kept for a period of time as required by local regulations. This is trial version www.adultpdf.com Implementing the research project 81 A common problem in research is the tendency of investigators to collect many data, much more than they can analyse or publish. This can result in an excessively large database and increase the chance of inaccuracy. Limiting the data to be collected to the essential data for the study, and eliminating redundancies, will enhance the accuracy of the study. A general advice to investigators is to be parsimonious (not to expand more than is necessary). Quality assurance and quality control A system for quality assurance must be implemented to ensure that the study is performed and the data are generated, recorded and reported in compliance with the protocol, good clinical practice and national regulations. In clinical trials, all sites and all data and documents must be available for verification. All observations and findings should be verifiable in order to ensure the credibility of data and to ensure that the conclusions presented are derived correctly from the raw data. Quality assurance is carried out with the following objectives: to ensure that no data are missing and to ensure that data are precise and accurate. Missing data will introduce a problem in the analysis of results, whether the data are missing because the measurement was not made or was not recorded. A special type of missing data is loss to follow-up. Loss to follow-up will decrease the number of subjects. Generally, when sample size is estimated, a provision is made for an estimated percentage of loss to follow-up. But this does not completely solve the problem. Loss to follow-up may introduce a bias in the study and discredit its conclusions. The subjects lost to follow-up may be different from the subjects who continued in the study. For example, subjects who develop serious side-effects, complications or even die may be disproportionately represented in the loss to follow-up. Monitoring during the study can help in reducing the problem of missing data. A computer program can help during data entry to ensure that all data are entered. The computer program will flag missing and out-of-range values. Inaccurate and imprecise data are a worse problem than missing data, because they may not be discovered after the fact. Only a systematic quality control programme will avoid the problem. Reliability of measurements is an important component of quality assurance. To test for intra-observer reliability, a common way is to do the measurements twice (test–retest reliability). The results obtained from the first test are then correlated with the second test. To test for inter-observer reliability, a common way is to have the same measurements done by two observers. The results obtained from the first test are then correlated with the second test. This is trial version www.adultpdf.com [...]... conclusions are anticipated during the research design Dummy cross-tabulations that will allow the conclusions to be made are developed and left empty to be filled when the data are available 8 .4 Calculations Numerical data can be summarized by calculating their central tendency and variability, by calculating percentage and proportions, and by calculating ratios and rates Computer software programmes have facilitated... in the sample indicated by the standard deviation For example, the standard error of a mean is calculated as the standard deviation divided by the square root of the number of observations This is trial version www.adultpdf.com 94 A practical guide for health researchers By itself, the standard error may have a limited meaning, but it can be used to produce a confidence interval, which has a useful... care Normally the care of animals should be under the supervision of veterinarians having experience in laboratory animal science The avoidance or minimization of discomfort, distress or pain to the animal is an ethical imperative Procedures with animals that may cause more than momentary or minimal pain or distress should be performed with appropriate sedation, analgesia, or anaesthesia in accordance... received and who did not receive prenatal care The accepted convention in analytical cross-tabulations is to put the categories of the dependent variable (birth weight) as column headings, and the independent variable (prenatal care) as row headings The totals are also put for the columns and for the rows If percentages are used, they should add up to a total of 100% Analytical cross-tabulations may focus... information may be lost by being too summarized If in doubt, classes are better initially chosen to be small rather than large Small ones can be easily amalgamated to form larger ones if needed Classes must also be mutually exclusive For example, if we tabulate data about the diastolic blood pressure, we may make the classes as 70 to 79, and 80 to 89, not 70 to 80 and 80 to 90 Cross-tabulation tables... practice of drug companies of paying a fee to This is trial version www.adultpdf.com 86 A practical guide for health researchers the investigator for every patient participating in a clinical trial Local research ethics committees should have the authority to audit the implementation of the research, and to contact research subjects 7.11 .4 Fiscal honesty Research programmes and projects are commonly supported... Graunt with a massive but unorganized mass of information about the ongoing drama of birth and death occurring all around him Graunt made a big effort to organize the data in a way that was probably inspired by his techniques for tracking his shop inventory He devised tables that could be easily updated He took great pains to reduce several confused volumes of information into tables and This is trial... more than one peak 8.6 Correlation In the context of correlation, data are classified as either independent or dependent variables Independent or input variables ordinarily have values that are autonomous of the dependent or outcome variables Because independent variables precede dependent This is trial version www.adultpdf.com 92 A practical guide for health researchers variables, they often are called... investigator to assign particular treatments to different subjects The decision is made by random allocation Deliberate scientific fraud is ethically unforgivable Fraud involves deliberate deception and may take the form of fabricating data, inventing patients, or manipulating data to provide a desired answer The pressure to “publish or perish” in academic institutions may be a factor, as well as the practice... www.adultpdf.com 88 A practical guide for health researchers succinct paragraphs He was able to compare changes in mortality tables over the years Graunt published the summary of his work, entitled “Natural and political observations made upon the bills of mortality”, in 1662 His book immediately attracted the attention of government leaders and prominent private citizens The story goes that King Charles . companies of paying a fee to This is trial version www.adultpdf.com 86 A practical guide for health researchers the investigator for every patient participating in a clinical trial. Local research. monitoring: data management (record keeping and data handling) and data quality (quality assurance and quality control). Record-keeping and handling of data All steps involved in data management should. vary and may all be meaningful in the presentation of the data. This is trial version www.adultpdf.com 90 A practical guide for health researchers Variability In addition to knowing the mean